Dataset generation Model training Verification

Output report - 2021-11-12 11:10

This is an automatic report summarizing the results for the model created by the interface for Machine Learning (iML)

Step 1: Dataset generation


Pairplot matrix of input data

Pairplot matrix of output data

Correlation matrix of output data

Step 2: Model training


This section shows various plots that summarize the training step

test_split_proportion0.8
normalization_scaler_typeMinMaxScaler
neural_network_structure
input_layer
typeinput_layer
activity_regularizer
kernel_regularizer
bias_regularizer
activation
input_shape
hidden_layer
typehidden_layer
neurons100
activationrelu
activity_regularizer
kernel_regularizer
bias_regularizer
hidden_layer_b
typehidden_layer
neurons50
activationrelu
activity_regularizer
kernel_regularizer
bias_regularizer
hidden_layer_c
typehidden_layer
neurons25
activationrelu
activity_regularizer
kernel_regularizer
bias_regularizer
hidden_layer_d
typehidden_layer
neurons12
activationrelu
activity_regularizer
kernel_regularizer
bias_regularizer
output_layer
typeoutput_layer
activationlinear
activity_regularizer
kernel_regularizer
bias_regularizer
neurons
training_options
epochs1500
batch_size125
gradient_descent_options
metrics
  • mean_squared_error
lossmse
test_split_random
use_custom_loss
save_model
criteria

Training plots

Step 3: Verification


Verification with testing dataset

Ground truth vs ML plots - Trained variables

Ground truth vs ML plots - Pairplots